Context: Major surgery induces a catabolic state resulting in a net loss of body protein.
Objectives: Our objective was to compare protein metabolism before and after surgery in nondiabetic patients with and without preoperative insulin resistance (IR). It was hypothesized that the anabolic response to feeding would be significantly impaired in those patients with preoperative insulin resistance.
Design: A hyperinsulinemic-euglycemic clamp has been used to identify two groups of patients: IR and insulin sensitive (IS). A tracer kinetics technique has been used to evaluate the metabolic response to food intake in both groups.
Setting: Patients undergoing cardiopulmonary bypass participated.
Patients or Other Participants: Ten IS patients and 10 IR patients were enrolled in the study.
Intervention: After an overnight fasting, a 3-h infusion of a solution composed of 20% glucose and of amino acids at a rate of 0.67 and 0.44 kcal/kg · h, respectively, was started in each group. Phenylalanine kinetics were studied at the end of fasting and feeding.
Main Outcome Measure: Effect of feeding on protein balance before and after surgery was evaluated. Protein balance has been measured as the net difference of protein breakdown minus protein synthesis.
Results: Protein balance increase after postoperative feeding was blunted only in the IR group. In contrast, in the IS group, the postoperative anabolic effect of feeding was the same as before surgery.
Conclusions: These findings propose a link between insulin resistance and protein metabolism. When non-IR patients are fed, a significant anabolic effect in the postoperative period is demonstrated. In contrast, IR patients are less able to use feeding for synthetic purposes.
Major surgery induces a catabolic state resulting in a net loss of body protein (1, 2), and up to 1500 g of lean tissue can be lost after major abdominal surgery (3, 4). Whole protein breakdown increases, thus mobilizing amino acids for synthesis of new proteins involved in tissue healing and for synthesis of acute-phase proteins in the liver (5). Because proteins represent structural and functional components, erosion of lean body mass may result in prolonged convalescence and increased morbidity (6, 7).
In addition to enhanced protein catabolism, surgery induces a state of insulin resistance (IR) characterized by an increased hepatic production of glucose and reduced peripheral utilization (8). IR is defined as a state in which physiologic concentrations of insulin are insufficient to prevent hyperglycemia (9). It has been reported that this postoperative state of IR lasts several weeks after which insulin sensitivity (IS) returns to normality (10).
The association between IR and postoperative protein catabolism has been demonstrated in a study in diabetic patients who underwent major abdominal surgery where postoperative whole-body protein breakdown was significantly greater than in nondiabetic patients (11). This would imply that preoperative IR impacts on intermediate postoperative metabolism and the degree of protein catabolism. In another study, in patients undergoing cardiopulmonary bypass (CPB) with normal preoperative IS, perioperative blood glucose concentration did not change. This same group did not require any insulin infusion throughout surgery. In contrast, those patients with decreased preoperative IS became hyperglycemic despite large doses of insulin (12).
An association between IR and protein breakdown with a blunted anabolic response to food intake has also been demonstrated in different settings such as aging (13) and obesity (9). At the present time, it is not known whether a preoperative state of IR, in nondiabetic subjects, would have an impact on the degree of postoperative protein catabolism.
This present study was therefore set up in patients scheduled for elective CPB, first to identify those subjects who had IR before surgery by using a hyperinsulinemic-euglycemic clamp, and second, to evaluate in these subjects the metabolic response to food intake before and after surgery using tracer kinetics. It was hypothesized that the anabolic response to feeding would be significantly impaired in those patients with preoperative IR and not in those who were insulin sensitive.
Patients and Methods
The study protocol was reviewed and approved by the ethics committee of “Ospedali Riuniti di Bergamo.” The purpose and potential risks of the study were explained to all patients in the preoperative clinic two weeks before surgery, and informed written consent was obtained from each subject.
Exclusion criteria were hepatic dysfunction (alanine aminotransferase >24 U/liter), renal dysfunction (creatinine >1.2 mg/dl), diabetes type 1 or 2 (glycated hemoglobin >6%), emergency surgery, recent systemic infection, use of corticosteroids, left ventricular ejection fraction higher than 40%, preoperative use of inotropic drugs, or intraaortic balloon pump.
Thirty-three consecutive patients were enrolled in the study and underwent the hyperinsulinemic-euglycemic clamp to be screened for IR. Once the number of 10 IS patients was achieved, the enrollment continued until the number of 10 IR patients was reached. Twenty-three IS patients were identified, but the last 13 were excluded from the study because the number of ten IS was already obtained.
As shown in the study outline (Supplemental Fig. 1, published on The Endocrine Society's Journals Online web site at http://jcem.endojournals.org), patients were admitted to the cardiothoracic ward 2 d before surgery to undergo the clamp after which they were assigned to two groups: the insulin-sensitive group (group IS) and the insulin-resistant group (group IR). On the day before surgery, a 6-h fast and fed tracer kinetics study was performed to measure whole-body protein metabolism. The clamp was repeated on the first day after surgery followed by the tracer study on the second postoperative day.
The clamp was performed the morning after an overnight fast according to a modified De Fronzo technique (14). An iv cannula was inserted in an antecubital vein and a primed infusion of biosynthetic regular insulin (Actrapid Novo Nordisk A/S, Bagsvaerd, Denmark) was started and maintained for at least 3.5 h at a rate of 40 mU/m2 · min to achieve plasma insulin concentrations between 500 and 600 pmol/liter. Another cannula was inserted in the contralateral radial artery for blood sampling. Arterial blood glucose was measured every 10 min, and glucose level was maintained at 5.5 mmol/liter by adjusting it with 20% dextrose that was infused together with insulin. Patients with a glucose uptake below 5.5 mg/kg · min were defined as insulin resistant (group IR), whereas patients with a glucose uptake of at least 5.5 mg/kg · min were defined as insulin sensitive (group IS) (14, 15) .
Blood samples for the analysis of serum insulin, cortisol, and C-reactive protein (CRP) were collected before clamp and at the end of the clamp before stopping insulin infusion.
Measurement of protein metabolism
The isotopes l-[ring-2H5]phenylalanine (99 atom percent excess), l-[ring-2H2]tyrosine (99 atom percent excess), and l-[ring-2H4]tyrosine (99 atom percent excess) were purchased from Spectra 2000-CIL (Rome, Italy). The chemical, isotopic, and optical purity of these compounds was confirmed before use. Solutions were prepared under sterile precautions in the hospital pharmacy and were shown to be bacteria and pyrogen free before use.
Whole-body protein kinetics was performed after an overnight fast and during the morning hours of the day before surgery and 2 d after surgery. The metabolic study outline is illustrated in Supplemental Fig. 2. After baseline blood sampling to measure isotope enrichment, a primed infusion of l-[ring-2H5]phenylalanine at 4.2 μmol/kg · h and l-[ring-2H2]tyrosine at 1.2 μmol/kg · h was started. After 180 min of the tracer infusion (fasted state), five blood samples were collected at 5-min intervals to determine tracer enrichments. A solution of 20% glucose and of amino acids (Isoselect 8%; Bieffe Medital S.p.A., Grosotto, Italy) was then started at a rate of 0.67 kcal/kg · h (2.1 ml/kg · h) and 0.44 kcal/kg · h (0.6 ml/kg · h), respectively, for 180 min, corresponding to 1.1 kcal/kg · h. At the end of the infusions, five blood samples were taken again for tracer measurement. Plasma concentrations of glucose and insulin were collected at 180 and 380 min into the isotopic infusion. Each blood sample was immediately transferred to a heparinized tube and centrifuged at 4 C (3000 rpm for 15 min), and the supernatant stored at −70 C until analysis.
Analysis of samples
Plasma enrichment levels of l-[ring-2H5]phenylalanine, l-[ring-2H2]tyrosine, and l-[ring-2H4]tyrosine were determined by gas chromatography mass spectrometry.
Plasma glucose was measured by a glucose oxidase enzymatic method [Roche/Hitachi 904/911 (Naples, Italy) modified glucose oxidase-piroxidase 4-amino antipyrine] with a coefficient of variation (CV) of 1.8% at 6.83 mmol/liter and 1.9% at 13.8 mmol/liter. Plasma cortisol was measured using the Ciba Corning ACS 180 automated immunoassay (Ciba Corning Diagnostic, East Walpole, MA). Serum insulin was measured by a solid-phase, two-site chemiluminescent immunometric assay (Immulite/Immulite 1000, Milan, Italy) with a CV of 6.4% at 7.39 μIU/ml and 5.3% at 300 μIU/ml. CRP was measured using a laser nephelometric technique (BN 100; Medgenix Diagnostics, Fleurus, Belgium).
Calculations of phenylalanine kinetics
Isotopic plateau (CV < 7%) was observed during the fasted (180–200 min) and the fed states (380–400 min). Mean values of isotopic enrichment values at each plateau were used for all calculations of amino acid kinetics.
Whole-body phenylalanine flux and tyrosine flux were calculated using the equations described by Thompson et al. (16) outlined below.
Infusion and flux rate units are micromoles per kilogram per hour. The rates of flux (Q) of Phe and Tyr (measured) were obtained from isotope dilution according to equation 1: where i is the rate of tracer infusion and Einfusate and Eplasma correspond to the enrichments of infusate and plasma amino acids, respectively. Phenylalanine flux (Q) measures protein breakdown in the fasted state. Phenylalanine flux (Q) minus phenylalanine rate of infusion as from amino acids administration (phenylalanine concentration 0.27 g/liter) measures protein breakdown in the fed state. Protein balance has been calculated as protein synthesis minus protein breakdown. Conversion rate of Phe to Tyr (Qpt) was calculated as where Qt and Qp are the flux rates for [2H2]Tyr and labeled Phe, respectively; Et and Ep are the Tyr [2H4]Tyr and labeled Phe enrichments in plasma, respectively; and ip is the infusion rate of the Phe tracer. The rate of whole-body incorporation of Phe into proteins, Sp, was calculated as the difference between Qp and Qpt.
Calculation of insulin effect on protein metabolism
To normalize the effect of insulin on protein breakdown, the percent increase of insulin after feeding was divided by the percentage in decrease of protein breakdown. This number reflects the percent increase in insulin levels necessary to determine the same unitary percent decrease of protein breakdown and was used here as an index of IR of protein metabolism (IRPM). The rationale of this method relies on previous studies that have shown that the stepwise effect of insulin of whole-body protein degradation is linear within the physiological range of insulin concentrations in both healthy subjects and insulin-resistant diabetics (17, 18). Nonetheless, the slope of regression lines was significantly different in controls and diabetics, suggesting that the ratio between protein kinetics and insulin levels is an index of insulin sensitivity on protein metabolism (17, 18).
Anesthesia and surgical care
Anesthesia was induced with propofol 1 mg/kg and remifentanil at 0.3 μg/kg · min, and muscle relaxation was achieved with vecuronium 0.1 mg/kg. Anesthesia was maintained with a continuous infusion of remifentanil at 0.4 μg/kg · min and propofol at 70 μg/kg · min and titrated to keep the bispectral index between 40 and 60. Continuous monitoring included electrocardiogram with computerized ST segment analysis of leads II and V5, oxygen saturation, invasive arterial blood pressure, cardiac output and cardiac index, mixed venous oxygen saturation, and esophageal temperature. Immediately before CPB, heparin 300 U/kg was administered. The activated coagulation time was maintained longer than 400 sec. Hypothermic CPB (to a lowest temperature of 32 C) was achieved with a membrane oxygenator and crystalloid prime (2.0 liters lactated Ringer's solution, 50 mEq sodium bicarbonate, 10,000 U heparin, and 12.5 g mannitol). Nonpulsatile flow was maintained from 2.4–2.8 liters/min · m2. Myocardial protection was with antegrade and/or retrograde hyperkalemic blood cardioplegia. Separation from CPB was facilitated with iv inotropic and/or vasoactive drugs. Anticoagulation was reversed with protamine 4 mg/kg.
After completion of surgery, patients were transferred to the intensive care unit, and patients were extubated clinically adequate. Postoperative pain relief was achieved with continuous iv infusion of morphine sulfate at 0.02 mg/kg · h and proparacetamol at 1 g every 8 h.
Protein and calorie intake (kilocalories) was measured during the day of the hyperinsulinemic-euglycemic clamp.
Sample size and statistical analysis
This study was designed to show the differences of protein metabolism between IS and IR patients. The main outcome was the difference in protein balance between the fasted and fed state. Assuming a difference of at least 1.2 μg/kg · min with a sd of 0.9 μg/kg · min between two groups, 10 patients in each group were required for an 80% power and 0.05% significance.
All quantitative variables are summarized as mean and sd in the tables and as mean and se in the figures. The qualitative variables are summarized as frequency and percentage. The results were reported separately for each of two groups (IS and IR).
Shapiro-Wilk test was performed to verify normal distribution.
Student's t test for unpaired data was applied for assessing the comparison of the quantitative variables between two groups (IR vs. IS). Fisher's exact test was used for qualitative variables.
Effect of nutrition on protein breakdown, protein synthesis, and protein balance was evaluated as the difference of the fed minus the fasted state for each of these variables. The resulting variables [change in protein breakdown (ΔPBreak), synthesis (ΔS), and balance (ΔPBal)] were analyzed using different linear mixed models for repeated data. The mixed model is a powerful method for analyzing data from longitudinal studies, in which there are multiple measurements on each subject (19). This approach allows explicit modeling of the within-person and between-person variation in the outcome while taking into account the correlation between repeated measurements on the same individual.
The linear mixed model for repeated measurements was used to regress pre- and postsurgery measures on the fixed-effect factors assuming unstructured covariance matrix.
The same model was applied on cortisol and CRP variables.
The linear mixed model for repeated measurements was also used to evaluate the effect of each factor (group, nutrition, and surgery) and their interaction (group × surgery, nutrition × surgery, and group × nutrition) on circulating levels of blood glucose and serum insulin.
In all models, a priori contrasts were used to compare means of different parameters between IR and IS patients at different time points or between pre and postsurgery values in either IR or IS group.
The correlations between postsurgery glucose uptake and protein balance was reported graphically as a scattergram and evaluated by a Spearman correlation coefficient (ρ).
All statistical analyses were performed using SPSS software version 11.0 (SPSS Inc., Chicago, IL).
There were no differences between the two groups with regard to the demographic characteristics and duration of surgery, CPB, and aortic cross-clamp. Body mass index was similar between the two groups. Normal-weight, overweight, and obese subjects were equally distributed among the two groups. Both groups received the same caloric intake during the days of the clamp (Table 1). The median visual-analogic score at rest on the first postoperative day was 2 (range 0–3) in the IS group and 2 (range 0–3) in the IR group. On the second postoperative day, the median visual-analogic score at rest was 1 (range 0–2) in the IS group and 1 (range 0–2) in the IR group.
In all experiments, isotopic tracer plateau enrichments were achieved during both fasted and fed states (mean CV = 7%), allowing the use of the steady-state equation. All raw data relative to net protein breakdown, synthesis, and balance are presented in Table 2.
To determine the effect of feeding on all aspects of protein kinetics, the difference between fasted and fed states of the two study groups were analyzed before and after surgery and presented in Table 3.
Feeding decreased the ΔPBreak and the ΔS in both groups, and this was similar before and after surgery. Feeding increased the ΔPBal in both groups before and after surgery (P = 0.036); however, after surgery, it was significantly less in the IR group (P = 0.009). Although the fall in ΔPBreak after feeding occurred in all groups before and after surgery, this decrease was significantly less in the IR group (P < 0.001). Similarly, the increase in ΔPBal after feeding was blunted after surgery only in IR group (0.009).
Circulating concentrations of insulin and glucose (Table 4)
Concentrations of serum insulin and blood glucose are represented in Table 4. Feeding increased insulin levels independently of surgery and group (P < 0.001). This effect was significantly greater in the IR patients (P < 0.001) before and after surgery (P < 0.001). Feeding had no effect on blood glucose concentrations. Surgery increased insulin levels independently of feeding and groups (P < 0.001). This effect was greater in the IR patients (P = 0.05). Surgery increased blood glucose levels (P < 0.001) independently of feeding and groups. This increase was higher in the IR group (P < 0.001) and after feeding (P = 0.030).
Insulin levels were lower in the IS group compared with the IR group independently of feeding and surgery (P < 0.001). The difference in insulin levels between groups increased after surgery (P = 0.05) and feeding (P < 0.001). Blood glucose levels were greater in the IR patients independently of feeding and surgery. The difference in glucose levels between groups increased after surgery (P < 0.001) and feeding (P = 0.03).
Insulin resistance of protein metabolism
Changes in IRPM, estimated by measuring the ratio of the percent increase after feeding of serum insulin levels over percent decrease of protein breakdown are illustrated in Supplemental Fig. 3A. IRPM increased after surgery independently of the group studied (P < 0.001). Interaction of surgery × group was not statistically significant (P = 0.44). Percent increase in insulin levels and percent decrease in protein breakdown rate after feeding are illustrated in Supplemental Fig. 3, B and C, respectively. Insulin percent increase after feeding doubled in the postoperative period in IS, whereas it did not change in IR (P < 0.05). Protein breakdown percent decrease after feeding remained the same after surgery in IS, whereas it was halved in IR (P < 0.05).
The results of the clamp are presented in Supplemental Fig. 4. The glycemic goal of 5.5 mmol/liter was reached before and after surgery in both groups, within a similar time interval (136.7 ± 21.6 min for IS and 127.9 ± 17.7 min for IR before surgery and 130.2 ± 26.6 min for IS and 125.6 ± 25.4 min for IR after surgery). Glucose uptake decreased after surgery in both groups (P = 0.005 for IS group and P = 0.037 for IR group), and the rate of decrease is comparable among groups (P = 0.716).
A correlation between glucose uptake and protein balance after surgery was calculated by including all patients regardless of the group. A positive correlation was observed after feeding (ρ= 0.603; P = 0.002) as shown in Supplemental Fig. 5.
CRP and cortisol (Table 5)
IR patients had a higher CRP concentration before surgery compared with IS patients. After surgery, CRP increased in both groups, and no differences were observed among groups in CRP levels. Plasma cortisol levels increased after surgery and were comparable among groups before and after surgery.
The main finding of this study is that the anabolic effect of glucose and amino acids was blunted after surgery in patients with preoperative IR. In contrast, postoperative protein balance after feeding in patients with preoperative IS remained the same as before surgery.
Surgery is associated with a state of IR, and this represents one of the key mechanisms by which more than a few common surgical complications are triggered (10). Several studies have demonstrated a positive correlation between hyperglycemia and/or perioperative IR and complications in elective surgery (20, 21).
There have been attempts to attenuate postoperative IR to improve postoperative protein economy and to decrease postoperative complications. For example, it has been shown that perioperative infusion of insulin aimed at maintaining tight euglycemia (57 ± 11 IU/24 h to achieve a mean blood glucose of 5.8 ± 0.4 mmol/liter) improved muscle protein synthesis (22). Similarly, preoperative high oral carbohydrate loading (125 mg/ml) vs. low carbohydrate loading (25 mg/ml) showed better postoperative whole-body protein balance and a better suppressive effect of insulin on endogenous glucose release (23). These results put in evidence a link between IR and protein metabolism and are in agreement with our finding of a positive correlation between postoperative glucose uptake and postoperative protein balance after feeding (Supplemental Fig. 5).
Although the two previously cited studies concentrated on treating postoperative IR without discriminating whether patients were insulin resistant before surgery, the present study also identified a correlation between preoperative IR and postoperative protein metabolism.
This kind of correlation has already been identified years ago by our group in diabetic patients (11). In that study, diabetic patients undergoing colon surgery had an increased postoperative oxidative protein loss and protein catabolism compared with nondiabetic patients. These findings lend support to the hypothesis that postoperative protein catabolism degree is not the same for all the patients. This could depend on their basal preoperative metabolic conditions.
In the present study, IRPM increased after surgery in both groups and preoperative and postoperative IRPM were not different among groups. However, the reasons for the postoperative increase were different according to group appartenance.
We have used the percent increase of insulin instead of absolute values because increased basal insulin levels do not eliminate the possibility of an inhibition of insulin secretion after a meal. In fact, after surgery, diabetic patients show higher postoperative insulin levels but also higher blood glucose levels because they are not able to augment insulin secretion normally after a meal (24). We have used percent decrease in protein breakdown rather that the absolute values because we wanted to know protein breakdown variation in relation to the active insulin, which is the one secreted after the stimulus.
In Supplemental Fig. 3A, the variation of IRPM before and after surgery is illustrated. IR increased after surgery to the same extent in both groups. However, the reasons for the postoperative increase were different. In the IS group, it was due to the higher percent increase of insulin after nutrition, whereas the decrease in protein breakdown did not change (Supplemental Fig. 3, B and C). On the contrary, in the IR group, insulin percent increase after feeding was minimal, but protein breakdown reduction was less pronounced compared with before surgery (Supplemental Fig. 3, B and C). In the IS group, the ratio increased because the numerator increased, whereas in IR group, it increased because the denominator decreased. We do not know why postoperative insulin secretion did not increase after the meal as before surgery in the IR group. This group displayed, before and after surgery, a basal hyperinsulinemia that probably should not be interpreted as evidence that B cell function is no longer impaired and that IR is the only important feature during the recovery stage from tissue injury in these patients.
Perioperative trends in insulin levels are not well characterized. In some studies, serum concentration of insulin after trauma and surgery is decreased due to impaired insulin secretion and degradation, whereas in others, the insulin levels increase due to decreased peripheral utilization of glucose and increased hepatic production, thus raising blood glucose and insulin concentrations (25,–,27). This lack of homogeneous results could also be due to lack of preoperative stratification of patients in IR and IS.
The present study may also shed some light on the debate around glucose control in the postoperative intensive care patient emerged after the Leuven study (28). Other studies were not able to find the same results or even found opposite results (29). The lack of stratification of patients in IR or IS before surgery could have contributed to the contradictory results.
It is of interest to observe that in those patients with preoperative IS, there was no change in postoperative protein balance (Table 2). These findings are in net contrast with all previous studies conducted in patients undergoing major surgery where the decrease in postoperative protein balance varied from 5–25% (5, 30). One possible explanation of this discrepancy could be the lack of preoperative stratification of the surgical patients in IR and IS, resulting therefore in different metabolic outcome.
A limitation of this study is the lack of normalized protein kinetics according to fat-free mass (FFM). This could lead to different results because it is known that FFM has higher rate of protein turnover compared with fat mass. Thus, individuals with a higher percent body fat might appear to have blunted amino acid kinetics when normalized to body mass but yet not blunted if normalized to FFM.
However, in the present study, normal-weight, overweight, and obese subjects were equally distributed among the two groups. The main results do not change when the obese subjects are removed. The main result of this study is related to the comparison of between-groups variations observed within groups. There is a difference in protein balance after nutrition in the IR group before and after surgery (the increase in protein balance after nutrition is blunted after surgery compared with before), but this difference is not evident in the IS group. This means that the true difference is within the IR group, before and after surgery.
Higher circulating concentrations of glucose reported here in IR patients (Table 4) confirm the previous report of hyperglycemia only in IR patients after CPB (12). In addition, in the present study, postoperative IR for glucose metabolism increased in both groups, but in IS patients, glucose disposal never reached true insulin-resistant values (<5.5 mg/kg · min) (Supplemental Fig. 4). Another dissimilarity between the IR and IS groups was the difference in preoperative CRP values, with greater CRP in IR patients. This is in agreement with several studies that have reported elevated levels of CRP, white blood cell count, uric acid, and fibrinogen in people with IR or metabolic syndrome (31,–,33). This difference disappeared after surgery when an inflammatory state developed also in insulin-sensitive patients.
Plasma cortisol concentrations were not different between the two groups; hence, cortisol cannot explain the observed difference in protein metabolism among groups.
In conclusion, these findings propose a link between IR of glucose metabolism and IRPM. When IS patients are fed, a significant anabolic effect in the postoperative period is demonstrated. In contrast, IR patients are less able to use feeding for synthetic purposes. More studies are warranted to better characterize the role of IRPM on postoperative catabolism.
Received March 3, 2011.
Accepted August 12, 2011.
Copyright © 2011 by The Endocrine Society
from source: http://jcem.endojournals.org/content/96/11/E1789.full#fn-group-1
There statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose,treat,cure,or prevent any disease.